Glioblastoma multiforme (GBM) is one of the most lethal cancers. Radiotherapy prolongs survival only modestly, because the dose is limited by the tolerance of the normal brain tissues. In order to improve the results of radiotherapy we propose to increase the intrinsic radiosensitivity of GBM by selectively targeting ATM (Ataxia-Telangiectasia mutated), which is a key mediator of the DNA damage surveillance pathway in an irradiated cell. As proof of principle, we downregulated ATM expression in GBM cells and enhanced their radiosensitivity. In-vitro, the surviving fraction after 2 Gy (SF2) decreased, from >0.5 without antisense to 0.28-0.35 with antisense. It is notable that the average SF2 of cells isolated from GBMs, which are incurable by radiotherapy, is 0.5 whereas the average SF2 of cells isolated from anaplastic astrocytomas, which are curable by radiotherapy, is 0.34. In-vivo about half the tumors were cured, with a dose of irradiation that, without the antisense, cured a few. In order to increase the transduction efficiency of our genetic antisense vectors in GBM cells, we constructed an E1B-deleted replicating adenoviral vector (Adeno-E1BE -alphaATM) and successfully attenuated ATM protein expression in U-87 (p53 w.t.) and U-138 (p53 mut) GBM cells resulting in enhanced radiosensitivity in vitro. Interestingly, BIBA-Adeno-aATM enhanced the tumoricidal effects of the parent Adeno-E1B virus, even without irradiation. We further demonstrated the safety of Adeno-E1B -alphaATM in human umbilical vein endothelial cells and mouse astrocytes in vitro, and the mouse brain in vivo. Finally, we demonstrated that the human hexokinase II (hHKII) promoter is induced 15-fold in GBM cells when compared to expression in cultured normal neurons and astrocytes. Hypoxia and irradiation further induced the hHKII promoter. We now propose: (I) To investigate the role of ATM and its downstream targets in determining the radiosensitivity of GBM cells. The adenoviral vectors will be used as tools to down regulate ATM in molecularly well-chracterized GBM cells. (II) To further enhance the therapeutic benefit of Adeno-E1B cLATM virus by regulating the expression of the antisense ATM RNA under the control of a tumor-specific, hypoxia-sensitive, radio-inducible hHKII promoter. The infectivity/tropism of Adeno-EIBA- aATM virus to GBM cells will be increased by constructing vectors with adenoviral fiber mutation, F/K20. (III) To construct conditionally replicating antisense-ATM herpes simplex virus (HSV) vectors, expressing antisense-ATM under the control of the HKII promoter. We will also examine whether neural progenitor cells can be used to deliver the HSV antisense-ATM vector to tumor cells in the brain. (IV) To investigate the combined toxicity of the virus vectors, ATM attenuation and radiation therapy in cultured endothelial cells, astrocytes, oligodendrocytes, neurons and brain tissues in mice models.